Abstract Fractal Image Compression (FIC) problem is a combinatorial problem which has recently become one of the most promising encoding technologies in the generation of image compression. While Quantum Evolutionary Algorithm (QEA) is a novel optimization algorithm proposed for class of combinatorial optimization problems, it is not widely used in Fractal Image Compression problem yet. Using statistical information of range and... Fractal Image Compression (FIC) problem is a combinatorial problem which has recently become one of the most promising encoding technologies in the generation of image compression. While Quantum Evolutionary Algorithm (QEA) is a novel optimization algorithm proposed for class of combinatorial optimization problems, it is not widely used in Fractal Image Compression problem yet. Using statistical information of range and domain blocks, and a novel magnetic update operator, this paper proposes a new algorithm in solving FIC. The statistical information of domain and range blocks is used in the initialization step of QEA. In the proposed update operator the q-individuals are some magnetic particles applying attractive force to each other. The force two particles apply to each other depends on their fitness and their distance. The proposed algorithm is tested on several images and the experimental results show better performance for the proposed algorithm than QEA and GA. In comparison with the full search algorithm, the proposed algorithm reaches comparable results with much less computational complexity. Read More Read Less

Abstract Charts are widely used in technical and business documents as a graphical representation of numerical and qualitative data. We present a model-based method to automatically extract data carried by charts and convert them to XML format, thus making these data available for indexing, querying and analysis by common methods of textual data management. The proposed method includes several steps: 1)... Charts are widely used in technical and business documents as a graphical representation of numerical and qualitative data. We present a model-based method to automatically extract data carried by charts and convert them to XML format, thus making these data available for indexing, querying and analysis by common methods of textual data management. The proposed method includes several steps: 1) chart detection, 2) model-based classification by chart type and extraction of graphical components, 3) detection and recognition of textual components, 4) extraction of semantic relations between graphics and text. For testing purpose, a benchmark set was created with the XML/SWF Chart tool. By comparing the recovered data and the original data used for chart generation, we are able to evaluate our information extraction algorithm and confirm its validity. We also extensively tested each step of the proposed algorithm against existing approaches. Modelbased classification showed high accuracy, comparable to those of the best supervised learning methods. The proposed text detection algorithm leads to significant improvement in text recognition rate (up to 20 times better). Read More Read Less

Abstract Relationship discovery has found considerable interest recently due to the growing number of social network users and the increasing need to analyze social interactions. With the huge amount of user-generated content, photos, in particular, are gaining popularity in social networks. Setting the focus on photos is of valuable importance for extracting essential social relationships’ evidences about users. In this paper,... Relationship discovery has found considerable interest recently due to the growing number of social network users and the increasing need to analyze social interactions. With the huge amount of user-generated content, photos, in particular, are gaining popularity in social networks. Setting the focus on photos is of valuable importance for extracting essential social relationships’ evidences about users. In this paper, we present a rule-based approach to discover such relationships among users (e.g., colleagues, relatives, friends, etc.). We propose to automatically generate the set of basic rules through a crowdsourcing methodology using photos. The creation of a knowledge base (i.e. a photo dataset), founded on the wisdom of a large number of web users, is the primary objective that enables the extraction of a set of basic rules. Extending the set of basic rules is another important and essential objective for relationship discovery. Within this context, we provide a set of manually generated common sense rules that are relationship- type independent. In addition, we use data mining to extend the basic rules in order to obtain another set called the derived rules. These derived rules take into account the context of each user and are interestingly useful to identify relationship types yet unlabeled. Experiments results demonstrate the correctness and the efficiency of the generated sets of rules for relationship discovery. Read More Read Less